摘要
针对当前的行李物品安检系统通过人眼进行判别,检测效率较低、存在漏检的问题,提出了一种改进后的Faster-RCNN跨层检测网络,通过跨层连接网络层,采集多种角度下的危险物信息,将网络结构中提取的低层次特征和高层次特征相结合来训练网络结构,提升网络的复杂性,最后将本文所设计的网络结构在数据集上进行训练和检测实验。实验结果表明,对于多种类型目标物检测效果良好。
In view of the current baggage and article security inspection system,the detection efficiency is low,and missing detections are common,because it mainly depends on human eyes to make the judgement,an improved Faster-RCNN cross-layer detection network was proposed,the network layers across layers was connected,the dangerous objects information from multiple angles was collected,and the low-level features and high-level features extracted from the network structure to train the network structure to increase the complexity of the network.Finally,the network designed was tested on the data set.The experimental results show that the detection effect is good for multiple types of targets.
作者
郭鹏程
张文琪
李毅红
GUO Peng-cheng;ZHANG Wen-qi;LI Yi-hong(Information Center,The ShanXi Science and Technology Department,Taiyuan 030051,China;Shanxi Provincial Key Laboratory of Information Detection and Processing,North University of China,Taiyuan 030051,China)
出处
《科学技术与工程》
北大核心
2020年第33期13718-13724,共7页
Science Technology and Engineering
基金
国家自然科学基金(61801437)
山西省自然科学基金(201801D221206,201801D221207)。
关键词
X射线检测
卷积神经网络
跨层网络
目标检测
危险物品
X-ray detection
convolutional neural networks(CNN)
cross-layer network
target detection
dangerous goods
